The goal of this project is to describe how an extrastriate visual area, MT, generates response properties not present in its inputs. One novel aspect of this study is that the average input properties will be determined first from 2nd order properties of the MT cell, then stimuli tailored to match the inputs will be used to study how the inputs are combined and modified by the MT cell. A full mathematical description of the local-motion detectors making up each MT cell's receptive field will be used to determine which response properties can be accounted for by the cell's inputs and which properties require further processing. For response properties that do require higher-level processing, the computations underlying these properties will be explored by comparing the response properties with the spatial distribution of the inputs and the interactions between those inputs. Understanding how excitatory and inhibitory inputs to an extrastriate area are spatially distributed and combined should elucidate fundamental principles of information processing and cortical organization. The proposed interaction studies will explore integrative processes that might underlie such phenomena as hysteresis in motion perception, motion cooperativity, speed tuning, and sensitivity to continuous trajectories. The proposed studies on speed selectivity as a function of contrast may have implications for the kinds of automobile driving errors expected for medical conditions that reduce vision contrast, such as early cataracts. This laboratory has used similar stimuli and analytical approaches to study direction, disparity, and color processing in primary visual cortex, achieving most of the goals for the previous proposal, and has recently successfully applied these approaches to extrastriate area MT. The proposed studies, and the novel iterative mapping technique, are an extension of these studies to the next-higher stage of motion processing.